BACK
Stagehaus
Dropped!
The idea was straightforward: dealerships spend significant time and money arranging professional photography for their forecourt stock — booking studios, hiring photographers, coordinating logistics. Stagehause was designed to cut all of that out, using AI image generation to produce premium staged car visuals at a fraction of the cost and turnaround time.
Results during development were genuinely good. The output quality was strong enough to take seriously as a commercial product.
The decision to drop it came down to two honest assessments. First, the consistency problem: AI model performance was good on average but not reliable enough at the margins. For a B2B product charging dealerships — where a botched image of stock on a live listing is a real problem — the margin for error had to be near zero. It wasn't there yet. Second, the moat question: outside the brand, the website, and the quality of client relationships, the core mechanic is not difficult to replicate. Anyone with access to the same models could rebuild it.
The passion faded alongside those realisations.
The brand and website are potentially for sale. The concept is probably better suited to someone already deep in AI art or prompt engineering — a creative who could both solve the consistency problem and bring genuine craft to the output. If a sale doesn't materialise, the option to return with a co-founder from that background remains open.